Specific deletion of RIM4 only in Purkinje cells reproduces motor dysfunction. (A) Representative fluorescent in situ hybridization (RNAScope) images showing the density of RIM4 mRNA (green) and Nissl counterstain (blue) in cerebellar Purkinje cells, cortex and hippocampus in RIM4 wild-type (WT) and KO_PCP2 mice. Scale bars in scans = 20 µm. The dashed white lines indicate the border between the granule cells (GC) and the Purkinje cells (PCs). (B) Quantification of RIM4 mRNA density in the three cerebellar subregions and pyramidal neurons in the cortex and hippocampus in RIM4 wild-type and KO_PCP2 mice. Purkinje cells, granule cells, molecular layer interneurons (MLI), pyramidal neurons (PN), n = 4 mice per group, unpaired t-test. (C) Graph showing the development of body weight during the first 35 days after birth in RIM4 wild-type and KO_PCP2 mice. n = 5 mice per group, two-way ANOVA test with Bonferroni post hoc correction. (D) Onset and duration of episodes of motor impairment in RIM4 KO_PCP2 during 5 days of continuous video monitoring. WT n = 5, KO_PCP2n = 7 mice. (E) Diagram showing the susceptibility of RIM4 wild-type and KO_PCP2 mice to the induction of an episode of motor impairment by intraperitoneal administration of saline (NaCl), ethanol (EtOH, 2 g/kg mouse), or caffeine (Caff, 25 mg/kg mouse). WT n = 4–8, KO_PCP2n = 6–10 mice. (F) Bar graph showing the latency to fall off an accelerating rotarod (4 to 40 rpm in 300 s) for RIM4 wild-type and KO_PCP2 mice (11–15 weeks), WT n = 10, KO_PCP2n = 8 mice, unpaired t-test. (G) RIM4 wild-type and KO_PCP2 mice (13–17 weeks) were placed in a cage with a running wheel and monitored for four consecutive days, regular wheel (RW) consisting of all 38 rungs for 2 days and complex wheel (CX) consisting of 22 rungs (16 rungs were randomly selected and removed) for 2 days. Black boxes indicate time spent running in the wheel, black lines represent episodes of motor impairment. WT n = 8 RIM4, KO_PCP2n = 5 mice. (H) Graph of the average instantaneous running speed of the mice shown in G. One-way ANOVA tests with Sidak's post hoc correction. (I) Representative images of a sagittal brain slices stained with fluorescent Nissl green from RIM4 wild-type (left) and KO_PCP2 (right) mice. White lines delineate the measured areas, cortex (CX), cerebellum (CB) and hippocampus (HC) (left, scale bars = 1 mm). Bar graph showing the quantitative analysis of the area of the analysed brain regions, CX (WT 1 ± 0.03, KO_PCP2 0.93 ± 0.037), CB (WT 1 ± 0.02, KO_PCP2 0.78 ± 0.037) and HC (WT 1 ± 0.078, KO_PCP2 1.05 ± 0.044, areas normalized to mean of WT). n = 5 mice per group, unpaired t-test. (J) Representative images of cerebellar sections (vermis VI) stained with calbindin (magenta) and VGlut2 (green) in RIM4 wild-type and KO_PCP2 mice. Scale bars = 50 µm. Line profiles of average VGlut2 intensity were measured from the Purkinje cell soma to the top of the molecular layer. (K) VGlut2 fluorescence intensity was plotted against the distance from the Purkinje cell soma for RIM4 WT (black) and KO_PCP2 mice (grey). The average length of the molecular layer for both groups is shown below the data (percentiles, 25, 50 and 75, are indicated by ticks on the lines). WT n = 5, KO_PCP2n = 6 mice (6–8 weeks), Kolmogorov-Smirnov test. (L) Juxtacellular recordings of Purkinje cells were performed in the presence of blockers for excitatory (20 μM CNQX) and inhibitory (10 μM gabazine) postsynaptic currents. Left: Quantitative analysis of the mean spontaneous firing rate during baseline recordings, unpaired t-test. Middle: Example traces of Purkinje cell firing at baseline and after 1 mM caffeine application in RIM4 wild-type (black) and KO_PCP2 (grey) mice. Red boxes above the traces indicate the onset of individual spikes. Right: Time course of the firing rate before, during and after caffeine application. n mice/n cells, WT 20/297, KO_PCP2 15/212 for spontaneous spike analysis and WT 5/20, KO_PCP2 3/17 cells for caffeine application. (M) Average spontaneous action potentials recorded in whole cell current clamp mode at resting potential. (N) Comparison of intrinsic excitability properties obtained from whole cell current clamp recordings (n cells, WT 9, KO_PCP2 10, except for input resistance: 6/6). (O) Deep cerebellar nuclei (DCN) neurons were filled with the red fluorescent dye TMR-dextran (200 µM) as shown in the fluorescence image. Purkinje cell axons were stimulated in the white matter. Example traces of inhibitory postsynaptic currents in Purkinje cells of RIM4 wild-type (black) and KO_PCP2 (grey) mice, stimulus indicated by a red box. Quantitative analysis of the paired-pulse ratio (PPR) (WT 0.98 ± 0.03, KO_PCP2 1.00 ± 0.02). Scale bars = 100 µm image, 20 ms (horizontal) and 0.1 nA (vertical) traces. n mice/n cells, WT 11/18, KO_PCP2 6/13, unpaired t-test. (P) Parallel fibres (PFs) were electrically stimulated with inter-event intervals of 50 ms. Representative excitatory postsynaptic potential (EPSC) traces at PF-PC from wild-type (solid black line) and knockout (KO, dashed line) mice. Red boxes indicate stimulus. Bar graphs showing PPR for both synapse types in WT (black) and KO (white). n = mice/n = cells, WT 5/9, KO_PCP2 3/6 at PF-PC synapses; WT 3/8, unpaired t-test.